Multivariate Functional Network Connectivity for Disorders of Consciousness

Recent evidence suggests that healthy brain is organized on large-scale spatially distant brain regions, which are temporally synchronized. These regions are known as resting state networks (RSNs). The level of interaction among these functional entities has been studied in the so called functional network connectivity (FNC). FNC aims to quantify the level of interaction between pairs of RSNs, which commonly emerge at similar spatial scale. Nevertheless, the human brain is a complex functional structure which is partitioned into functional regions that emerge at multiple spatial scales. In this work, we propose a novel multivariate FNC strategy to study interactions among communities of RSNs, these communities may emerge at different spatial scales. For this, first a community or hyperedge detection strategy was used to conform groups of RSNs with a similar behavior. Following, a distance correlation measurement was employed to quantify the level of interaction between these communities. The proposed strategy was evaluated in the characterization of patients with disorders of consciousness, a highly challenging problem in the clinical setting. The results suggest that the proposed strategy may improve the capacity of characterization of these brain altered conditions.